On Thu, Apr 15, 2021 at 10:50 PM Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote:
> On Thu, Apr 15, 2021 at 2:26 PM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> > It's possible that that argument doesn't apply to the way SIGURG is used
> > in this patch, but I don't see a good reason to ignore the convention of
> > setting up the handler this way.
>
> Yeah, will fix. I don't think there is a bug here given the way
> latches use shared memory flags, but it might as well be consistent.
Here's a patch to change that. But... on second thoughts, and after
coming up with a commit message to explain the change, I'm not 100%
convinced it's worth committing. You can't get SIGURG without
explicitly asking for it (by setting maybe_sleeping), which makes it a
bit more like SIGALRM than SIGUSR2. I don't feel very strongly about
this though. What do you think?